{"paper_id":"c0a509d0-bd34-40cc-a279-797203a0eb56","body_text":"Over the last two decades, the prostaglandin-E 2  (PGE 2 ) receptor EP2 subtype has gained tremendous\nattention. Using\ngenetic knockout and pharmacological methods, the roles of EP2 have\nbeen delineated in various central nervous system (CNS) and peripheral\ndisease models. 1  EP2 is a Gα s -protein coupled receptor, which upon binding with endogenous\nligand PGE 2  activates adenylate cyclase resulting in the\nsynthesis of cAMP which promotes intracellular signaling via protein\nkinase A (PKA) or exchange protein activated by cAMP (Epac). 2  The PKA-mediated signaling is associated with\nneuroprotection and neuroplasticity, whereas Epac-signaling is associated\nwith neuroinflammation and neurodegeneration in the central nervous\nsystem. 2  Physiologically, the Gα s -coupled (cAMP) mediated signaling is also associated with\nsmooth muscle relaxation; thus EP2 acts as muscle relaxant. In the\nperiphery, EP2 receptor also promotes G-protein independent signaling\nthrough β-arrestin via c-Jun-N-terminal kinase (JNK) or extracellular\nsignal regulated kinase (ERK) leading to cancer proliferation, tumorigenesis,\nand metastasis. 2  Moreover, EP2 activation\nincreases IL-23 expression that causes T-cells to differentiate to\nTh17 (from Th0) effectors, leading to chronic inflammation, and increased\nrecruitment of neutrophils and macrophages to the injured site resulting\nin exacerbation of disease pathology. 3\nThe structural homology (amino acid identity) between the EP2,\nEP1, EP3, and EP4 receptors, which share the common endogenous ligand\nPGE 2  for their activation is limited to only 20–30%. 4  Among these, EP4 is also coupled with Gα s -protein and mediates the cAMP signaling like EP2 receptor.\nTherefore, EP2 and EP4 have similar functional roles in several diseases.\nInterestingly, EP4 also plays an opposing functional role in several\nother diseases (see below). The EP1 is Gq-coupled and regulates phosphoinositide\n3-kinases (PI3K) raising the cytosolic Ca 2+  levels, whereas\nEP3 is G i -coupled and inhibits the adenylate cyclase resulting\nin lowering the levels of cAMP. In addition to EP receptors, EP2 receptor\nhas structural similarity to extended family members of prostanoid\nreceptors such as DP1 (44%), IP (40%), and FP and TP (20%). 4  It is important to highlight that DP1 and IP\nreceptors are coupled to Gα s -protein. DP1 receptor\nupon binding with endogenous ligand PGD 2 , and IP receptor\nupon binding with endogenous ligand PGI 2  activate the adenylate\ncyclase resulting in synthesis of cAMP similar to EP2 receptor. This\ncAMP activity is involved in smooth muscle relaxation. 3  Functionally, DP1 receptors play similar roles as EP2 depending\non the disease phenotype, while the IP receptor role is mainly cardioprotective,\nmeaning inhibition of IP receptor would lead to adverse effects on\nthe cardiovascular system and other physiological functions. 5\nRecently, several studies have reported\nEP2 receptor as a key promoter\nof neuroinflammation in brain injury models and peripheral inflammatory\ndiseases suggesting EP2 inhibition with small molecules would be therapeutically\nbeneficial. 1  Some studies also indicate\nthat EP2 activation with agonist is therapeutically beneficial in\nhealing the bone from a fracture, 6  reducing\nintraocular pressure, 7  and anti-inflammatory\nand bronchodilatory effects in the lung, 8  reinforcing the idea that EP2 receptor is a novel therapeutic target\nfor drug discovery for a variety of diseases. However, there are also\nsome potential concerns about EP2 as a therapeutic target given its\nmechanism of action and “yin–yang” of the functional\nroles in physiological and disease-dependent pathological conditions.\nIn this Perspective, we evaluate the  Strengths, Weaknesses,\nOpportunities, and Potential Threats (SWOT)  of the EP2 as\na target for drug discovery, highlighting the strengths of proof-of-concept\nand efficacy studies in animal models, limitations associated with\nEP2 receptor targeting, and future studies needed to fulfill the knowledge\ngaps for clinical advancement of EP2 therapeutics. At the end, we\nalso briefly highlight SWOT analysis of the currently available lead\nEP2 antagonists for preclinical development and clinical trials.\n\nThe goal of this Perspective is not to review\nall the proof-of-concept\nstudies reported so far in various disease models exploring EP2 involvement,\nfor which the reader is directed to recently published review articles. 1 − 3  However, this Perspective’s purpose is to highlight the strengths\nand weaknesses of targeting EP2 receptor highlighting the opportunities\nand threats with studies that were conducted with scientific rigor\nusing in vitro and in vivo models, and where both genetic and pharmacological\napproaches provide cohesively strong support as described below.\nA recent study by Minhas et al. (2021) 9  shows that EP2 receptor expression is higher in aged immune cells\n(human and mouse macrophages) than in young cells. Aged myeloid cells\n(microglia and macrophages) heavily depend on balanced glucose levels.\nEP2 receptor activation in aged microglia and macrophages promotes\na microenvironment that converts glucose into glycogen, reducing the\nglucose flux and mitochondrial respiration and creating an energy\ndeficient state that drives a malign inflammatory state. 9  In aged mice, conditional deletion of EP2 from\nmyeloid cells or treatment with EP2 antagonists rejuvenates cellular\nbioenergetics, systemic and brain inflammatory states, synaptic plasticity,\nand spatial memory, and blocking peripheral myeloid cell EP2 signaling\nrestores cognition in aged mice, suggesting a role for EP2 receptor\nsignaling in promoting the youthfulness of immune functions. 9\nMicroglia (resident macrophages in the\nbrain) perform critical functions such as clearing misfolded proteins\nand invading pathogens and balance trophic factors that maintain normal\nneuronal function. In the Alzheimer’s disease (AD) brain, these\nbeneficial functions of microglia are impaired resulting in enhanced\nsynaptic and neuronal loss. EP2 engagement in microglia suppresses\nthe beneficial homeostatic functions of microglia, as a result the\nremoval (or phagocytosis) of amyloid-β (Aβ) plaques is\ninefficient. A study by Johansson et al. (2015) 10  showed that conditional deletion of EP2 in microglia in\na mouse model of AD restores chemotaxis, Aβ-clearance, regulation\nof inflammatory milieu, and regeneration of trophic factors, leading\nto prevention of loss of synaptic proteins and cognitive deficits.\nInterestingly, ablation of microglial EP2 signaling improved spatial\nmemory and increased presynaptic proteins in the APP-PS1 mouse model\nof AD. 10  Along the lines of these findings,\nearlier reports indicate that microglia isolated from EP2 global knockout\nmice show enhanced phagocytosis of Aβ, 11  and pharmacological antagonism of EP2 receptor with a small molecule\ninhibitor (C52,  Figure  2 ) enhances peritoneal macrophage mediated phagocytosis of Aβ 42 , 12  a key driver of AD. In an\nemerging study from our laboratory, chronic treatment of EP2 antagonist\nin 5×FAD mice starting at the prodromal stage (starting from\n3 months of age until they are 5 months) in drinking water showed\nreduced inflammatory mediators and gliosis in the cortex, 13  and this effect was only found in a two-hit\nmodel of 5×FAD (genetic 5×FAD mice were subjected to chronic\nbut mild LPS treatment for 2 months). 13  These studies conclude that EP2 is deleterious in age-related and\nAD conditions; therefore small molecule inhibitors (EP2 antagonists)\nmust be advanced for the treatment of age-related and Alzheimer’s\ndiseases.\nMicroglia (myeloid cells) play a key role in several\nother neurodegenerative\ndisease pathologies. Acute brain injuries due to status epilepticus\n(SE) and traumatic brain injury (TBI) create massive neuroinflammation\n(microgliosis, astrogliosis, and induction of cytokines, chemokines,\nand cyclooxygenase-2 (COX2)) in the brain, 2  which will exacerbate secondary neurodegenerative pathology leading\nto cognitive and behavioral deficits. Our laboratory investigated\nthe role of EP2 in SE models (mouse and rat), 14 − 18  and fluid percussion (TBI) injury in rats ( Figure  1 ). 19  Very recently, we (Varvel et al, 2021) 20  have shown that conditional ablation of EP2 from blood\nmonocytes or systemic EP2 antagonism with EP2 antagonist (TG6-10-1,  Figure  2 ) blocks monocyte entry to the mouse brain after SE, and post-SE\ntreatment with an EP2 antagonist remarkably prevents the breakdown\nof the blood–brain barrier (BBB), up-regulation of inflammatory\nmarkers, and neurodegeneration in the hippocampus of mice 3 days postrecovery\nfrom pilocarpine-induced SE. We have also reported that treatment\nwith another EP2 antagonist, TG11-77, reduces not only microgliosis\nin the hippocampus but also the cognitive deficits determined 8–18\ndays after recovery from pilocarpine-induced SE in mice. 21  In this later study, TG11-77 did not provide\nany neuroprotection in the hippocampus after SE injury, providing\na rationale for the hypothesis that anti-inflammatory efficacy is\nsufficient to modulate function of microglia to enhance cognitive\nfunction. These studies in conjunction with several others in rat\nand mouse models show that EP2 antagonism reduces the delayed mortality,\nexpression of neuroinflammatory mediators, and neurodegeneration in\nthe hippocampus, repairs the BBB, and prevents peripheral myeloid\ncell entry into the brain days following SE. This evidence strongly\nindicates that EP2 receptor is a suitable drug target for the development\nof therapeutic agents for treatment of the consequences of SE injury\n(event) and the ensuing cognitive deficits and potentially for the\nprevention of epileptogenesis after traumatic brain injuries. 19\n(A) Illustration of EP2 involvement in gliosis and subsequent\nneurodegenerative\npathologies such as epilepsy, memory and cognitive deficits, and Alzheimer’s\ndisease type dementia. (B) EP2 involvement in inflammation driven\ncancer proliferation, metastasis, and tumor development. EP2 inhibition\nwith a small molecule antagonist should be therapeutically beneficial.\nEP2 antagonists so far tested in animal proof-of-concept\nstudies.\nAmong these, AH 6809 is a dual antagonist of EP1 and EP2 receptors\nwith equal potency.\nCognitive impairments are common among the survivors\nof stroke,\nsepsis, and respiratory syncytial virus (RSV) and other severe acute\nrespiratory syndrome (SARS) virus infections. Lipopolysaccharide (LPS)\nhas been shown to mimic sepsis phenotypes. Systemic exposure to LPS\ninduces massive neuroinflammatory conditions and cognitive deficits\nafter recovery as recently shown by Jiang et al. (2020). 22  In this study, EP2 antagonist TG6-10-1 treatment,\nattenuated not only the massively up-regulated neuroinflammation (microgliosis\nand inflammatory markers such as IL-6, IL-1β, TNFα, COX2,\niNOS) in the hippocampus and loss of synaptic proteins (PSD-95) but\nalso depressive symptoms and memory impairment in mice. Since this\nLPS model did not (or does not) present any neurodegeneration phenotype\nin the hippocampus, we have not been able to confirm model independent\nneuroprotective efficacy by the EP2 antagonist in this model. Nonetheless,\nthese results support the hypothesis that neuroinflammation alone\nis sufficient to cause cognitive deficits in rodents; EP2 antagonists\nwith and without neuroprotective activity will offer cognitive improvements, 21  providing support toward clinical advancement\nof EP2 antagonist for several CNS diseases such as SE, AD, and aging-related\nillnesses. Please see  Table  1  for an overview of efficacy results described in this section.\nThe above studies indicate that EP2 receptor suppresses\nbeneficial\nfunctions of microglia and blocking myeloid EP2 signaling reduces\npathology in inflammatory neurodegenerative models. Another recent\nstudy by Liu et al. (2019) 23  reports that\nin mouse model of stroke, in which the initial ischemic event was\nfollowed by extended poststroke inflammatory response, EP2 knockdown\nfrom myeloid cells (Cd11bCre:EP2 lox/lox  mice) attenuated\nthe infiltration of macrophages (Cd11 + CD45 hi ) and neutrophils (CD45 + Ly6G hi ). Inducible\nglobal deletion of EP2 receptor in adult mice (ROSACreER;EP2 lox/lox ) also reduced the infiltration of myeloid cells to the brain and\nstroke severity in these mice. 23  EP2 expression\nis highly induced in neurons after ischemic injury, postnatal removal\nof neuronal EP2 in mice also reduced cerebral ischemic injury (infarct\nvolume) in a middle cerebral artery occlusion (MCAo) stroke model,\nsuggesting that EP2, irrespective of its cell origin, is involved\nin inflammatory brain damage and inhibition of EP2 signaling is protective\nafter ischemic stroke events. 23  Furthermore,\nthese findings were reproduced with a pharmacological treatment of\nbrain permeable EP2 antagonist (C52,  Figure  2 ) 4.5 and 24 h after MCAo injury to mice,\nwhere this antagonist reduced the infarct volume and improved the\nneurological score consistent with results found in ROSACreER;EP2 lox/lox  mice. 23  These results were\nrecapitulated by a similar study using another brain-penetrant EP2\nantagonist, TG6-10-1, which showed decreased neurological deficits\nand infarct volumes as well as down-regulated inflammatory cytokines\nin the brain in a transient (MCAo) mouse model ischemia 24  suggesting a novel strategy and strong rationale\nto develop therapeutic agents for treatment of stroke consequences\nby targeting EP2 receptor with small molecule antagonists.\nIt\nhas been shown that EP2 deletion (global knockout) reduces Aβ-load\nand oxidative stress in a mouse model of AD 25  extends the survival of mice and improves the motor strengths in\nan ALS model, 26  and reduces neurotoxicity\nin a model of Parkinson’s disease. 27  Several other studies showed that the PGE 2 /EP2 axis activates\nseveral innate immune pathways. 28 − 30  Multiple sclerosis is an inflammatory\nautoimmune disorder of the CNS. COX2, mPGES-1, and EP2 expression\nare elevated in patients with multiple sclerosis. COX2 deletion or\nCOX2 inhibition by celecoxib or EP2 inhibition by a dual EP2/EP1 antagonist\nAH6809 ( Figure  2 ) reduced\noligodendrocyte apoptosis, degree of demyelination and motor dysfunction\nin a cuprizone-induced model of multiple sclerosis. 31  It is very important to note that several of these central\nnervous system disorders are associated with cognitive impairment\nand motor disabilities. Given our consolidated findings that suggest\nthat EP2 driven neuroinflammation is strongly linked to cognitive\nand memory impairments and the fact that systemic administration of\nEP2 antagonist has attenuated cognitive and memory impairments and\nimproved neurologic score, we foresee that EP2 receptor must be explored\nas a drug target for several of these neurologic diseases ( Table  1 ).\nCOX2 is highly\nexpressed in a variety of cancers and exacerbates tumor aggressiveness\nthrough generation of precursor PGH 2  for the synthesis\nof PGE 2 . 36  Microsomal prostaglandin\nE synthase-1 (mPGES-1), the enzyme responsible for the last step of\nsynthesis of PGE 2  (from PGH 2 ), is also highly\ninduced in a variety of tumors. 37 , 38  EP2 receptor activation\n(by PGE 2 ) is associated with amplification of inflammation\nin the tumor microenvironment via induction of tumor promoting cytokines,\nchemokines, and growth factors. 2  There\nis a positive correlation between COX2, mPGES-1, and EP2 receptor\nexpression and inflammatory mediators that promote tumor proliferation,\nsurvival, migration, invasion, angiogenesis, and immune evasion in\nhuman gliomas. 34  The study from Qiu et\nal. (2019) 34  indicates that EP2 activation\ndrives human glioma cell (GBM) proliferation and invasion in cell\nculture models in vitro that overexpress COX2 (LN229, and SF767) and\noverproduce PGE 2 . An EP2 antagonist, TG6-10-1, blocked\nthe proliferation and invasion of these GBM cells, promoting apoptosis\nand cell cycle arrest. Moreover, in athymic nude mice that were inoculated\nsubcutaneously with COX2 overexpressing SF767 cells, oral treatment\nwith an EP2 antagonist for 4 weeks, significantly reduced the growth\nof subcutaneous tumors formed by SF767 cells, in which the average\ntumor burden (weight) was reduced by 63%. Additional experiments in\nthis study indicate that glioblastomas typically display angiogenesis\nhallmarks (determined by platelet endothelial cell adhesion molecule\n1 or CD31), which were increased by COX2 overexpression and then decreased\nby EP2 antagonist TG6-10-1 treatment indicating the role of EP2 in\nCOX2 driven angiogenesis of gliomas. Furthermore, 4 weeks of EP2 antagonist\ntreatment also suppressed orthotopic malignant gliomas from intracranially\ninjected luciferase labeled LN229 glioblastoma cells in nude mice. 34  These results are compelling to promote EP2\ninhibitors for treatment of glioblastoma multiforme when the compound\nTG6-10-1 or any other EP2 antagonist meets the requisite ADMET characteristics\nfor clinical development and clinical use.\nThe expression of\nEP2 receptor is very high in high-risk neuroblastoma (NB) in comparison\nto other PGE2 receptors. The expression of EP2 is elevated among the\nnonsurviving patients compared to surviving NB patients, indicating\nthat EP2 expression is coupled to poor survival of NB patients. 35  Similarly, EP2 receptor is expressed at higher\nlevels in NB patients with oncogenic MYCN gene amplification compared\nto NB patients with MYCN normal status suggesting a strong link to\nEP2 receptor in NB. MYCN amplification is the best characterized genetic\nmarker of high risk and chemoresistance in NB. 39  Interestingly, the other PGE 2  receptors (EP1,\nEP3, and EP4) showed inverse correlation with MYCN in NB. 39  Like in the glioblastoma study (above), the\nexpression of EP2 was correlated with several tumor promoting cytokines,\nchemokines, growth factors, and receptors including anaplastic lymphoma\nkinase receptor (ALK), brain derived neurotrophic factor (BDNF), chemokine\nligand-2 (CCL 2 ), chemokine receptor-2 (CCR2), colony-stimulating\nfactor-1 receptor (CSF1R), epidermal growth factor receptor (EGFR),\nand others. 35  When NB cell lines with various\nrisk factors (11q deletion (SK-N-AS), ALK mutation, MYCN amplification,\nP53 dysfunction, or KRAS mutation) were treated with EP2 agonists\n(PGE 2  or butaprost,  Figure  3 A), but not EP4 agonist (CAY10598,  Figure  3 B), they displayed induction\nof cAMP, and this cAMP induction was similar to the activity of an\nadenylyl cyclase activator forskolin, suggesting that EP2 is involved\nvia a Gα s -coupled mechanism in these cell lines.\nIn a key in vivo experiment to determine EP2 involvement in tumorigenesis,\n11q deleted NB cells (SK-N-AS), subjected to EP2 deletion by CRISPR–Cas9,\nor wild-type SK-N-AS cells were inoculated to athymic nude mice to\ngenerate high risk tumors with 11q deletion. The results show that\ntumors generated by EP2 deleted NB cells were significantly smaller\nin volume than the tumors generated by wild-type NB cells. Other than\ntumor volume, the mice that were given EP2 deleted NB cells were healthier\noverall, suggesting that EP2 is required for human high-risk NB cells\nto develop tumors. These findings were confirmed by multiple approaches\nincluding conditional deletion of EP2 and, importantly, by pharmacological\ntreatment with EP2 antagonist TG6-129 ( Figure  2  and see  Table  1 ), for 18 days in a SK-N-AS NB inoculated\nmouse xenograft model. It is interesting to note that systemic treatment\nof TG6-129 substantially decreased the proliferation of tumors formed\nby SK-N-AS cells in a dose-dependent manner, about 25% reduction with\na 10 mg/kg dose and 55% reduction with a 20 mg/kg dose for 18 days\nof once daily treatment. 35  These studies\nstrengthen the clinical development of small molecule EP2 antagonists\nand the EP2 receptor as a novel therapeutic target for a variety of\nmedically untreated cancers.\nEP2 and EP4 agonists used in the described studies.\nCOX2 is up-regulated in endometriotic tissue and eutopic endometrial\ntissue, which synthesize a high level of PGE 2  contributing\nto pathogenesis and exacerbation of endometriosis disease. 40 , 41  Inhibition of COX2 decreases survival, migration, and invasion of\nendometriotic cells that are associated with decreased PGE 2 . 41  There is a positive correlation between\nendometriosis induced vaginal hyperalgesia and the peritoneal fluid\nlevels of PGE 2 . 33 , 42  Moreover, EP2 receptor\nexpression is also very high in the uterus, 20  stromal cells in lesions, and mesothelial cells in the peritoneum. 33  Based on this COX2/PGE 2 /EP2 signaling,\nGreaves et al. (2017) 33  tested the role\nof EP2 in a mouse model of endometriosis monitoring the endometriosis\npain mediated behaviors (licking and exploratory activities) and mechanical\nwithdrawal from von Frey filament test. Treatment with EP2 antagonists,\nTG6-10-1 or PF-04418948 ( Figure  2 ) (10 mg/kg dose), resulted in statistically significant\nreversal of mechanical allodynia on the abdomen and hind-paw tests\n(see  Table  1 ). Interestingly,\noral administration of PF-04418948 displayed time-dependent effects\non mechanical withdrawal response in mice with endometriosis, with\nthe withdrawal threshold significantly lowered in both abdomen and\nhind-paw tests of endometriosis mice compared with the naive controls. 33  Although these results provide a strong impetus\nto advance an EP2 antagonist toward treatment of debilitating and\nmedically unaddressed endometriosis disease, the detailed molecular\nand phenotypic changes in the endometrium are not investigated in\nthis study to link it to the behavioral benefits. Moreover, the behavioral\nstudy was conducted with only  n  = 5 animals in each\ngroup; therefore, additional work with a larger number of animals\nis needed to enhance the strength of this work before the advancement\nof the EP2 target for endometriosis therapy.\n\nEarlier studies to delineate the function\nof EP2 receptor in models\nof stroke, AD, SE, PD, ALS, and innate immunity were mostly carried\nout using EP2 global knockout (EP –/– ) mice\nand in some cases using EP2 agonists (PGE 2 , butaprost,\nand CP-533,536,  Figure  3 ), because of the lack of EP2 antagonists until 2011–2012\nwhen highly characterized and selective EP2 antagonists were made\navailable from Pfizer 32  and Emory University 43  laboratories ( Figure  2 ). Subsequently, Amgen also reported a new\nclass of EP2 antagonists in 2015 for investigation with in vitro and\nin vivo models. 12\nIn a few studies,\nthe results from EP2 global knockout mice are\ndiscordant with pharmacological inhibition or conditional deletion\nof EP2 from adult mice. For example, in a MCAo model of stroke, EP2\nglobal deletion increased cerebral injury, with mice exhibiting impaired\nlearning and memory, 44  whereas conditional\ndeletion of EP2 in microglia or neurons of adult mice attenuated the\ncerebral injury, with the mice exhibiting normal learning and memory\nphenotypes. 23  Studies like these may create\na perplexing view among the pharmaceutical community for discovery\nand advancement of any therapeutic agent to clinical trials for this\ndisease. However, careful review of the confounding effects of EP2\nat the developmental stage (pre- and postnatal period) versus adult\nstage, as shown by Liu 23  offers a path\nforward. Several in vitro studies employing embryonic cortical or\nhippocampal neurons or hippocampal slice cultures from early postnatal\nbrain suggest that EP2 is neuroprotective when pharmacologically activated\nwith EP2 agonists or allosteric potentiators. 44 − 46  These studies\nwere conducted with insulting agents, glutamate, NMDA, and/or oxygen-glucose\ndeprivation. However, given the recent assessment of EP2 expression\nin neurons of the adult brain, which is low in comparison to embryonic\nneurons, those in vitro culture results need to be carefully interpreted\nprior to comparing with in vivo efficacy studies with EP2 deletion\nand pharmacological approaches.\nOne of the key homeostatic functions\nof activated microglia is\nphagocytosis to clear debris in the brain. An in vitro study using\nmicroglia isolated from EP2 deleted mice (P1–P3 neonates) showed\nenhanced phagocytosis of Aβ-peptides from AD brain sections,\ncompared to wild-type microglia. 11  However,\nwhen BV2 microglia cells that overexpress human EP2 receptors are\nused for phagocytosis of fluorescent latex microspheres, the presence\nof EP2 but not the activation or inhibition of EP2 (by selective EP2\nagonist or antagonist) caused an effect on the cells to phagocytose\nthese latex microspheres. 47  There are two\nmain differences in these studies; we are comparing the results of\nprimary microglia activity versus a transformed microglia like cell\nline (BV2) which may have confounding differences. 47  Moreover, the subject of phagocytosis is different among\nthe two studies (Aβ peptide vs fluorescent microspheres). Therefore,\nthese studies must be interpreted independently and must not be compared\nto one other to draw conclusions on the role of EP2 in phagocytosis\nin vivo. Nonetheless, several other in vitro studies using primary\nmicroglia and peritoneal macrophages are consistent with an antiphagocytic\nrole of EP2, where EP2 activation by PGE 2  suppressed macrophage\nmediated phagocytosis of anti-CD36, 48  which\nwas reversed by treatment with a nonselective EP2/EP1 antagonist,\nAH 6809 ( Figure  2 ).\nNagano et al. reported that EP2 receptor activation with PGE 2  dose-dependently reduced rat primary microglia mediated phagocytosis\nof amyloid-β 42 , which was reversed by treatment with\nAH 6809. 49  Similarly, mouse peritoneal\nmacrophage cells (IC21, ATCC TIB-186) when treated with EP2 antagonist,\nC52 ( Figure  2 ), showed\ndose-dependent increase in phagocytosis of Aβ-plaques present\nin brain slices from 18 month old Tg2576 mice (AD). 12  This effect of C52 (1 μM) was similar to that caused\nby 10 μg/mL Aβ-antibody in the same ex vivo assay, suggesting\nthat EP2 expressed in myeloid cells is strongly associated with phagocytosis\nand it can be modulated with activation and inhibition strategies.\nSimilarly, several in vitro studies from our laboratory using primary\nrat microglia or mouse BV2-microglia cells overexpressing human EP2\nreceptors indicate that EP2 activation with an agonist results in\nup-regulation of several proinflammatory cytokines and chemokines\n(IL-1β, IL-6 and CCl2) and down-regulation of cytokine TNF 47 , 50 , 51  suggesting that EP2 receptor\nhas a mixed impact on proinflammatory gene expression. Pretreatment\nof these cultures with selective EP2 antagonists reverse inflammatory\nmediators expression (i.e., down-regulation of IL-1β, IL-6 and\nCCl2 and up-regulation of TNF) suggesting that EP2 antagonism is not\ncompletely a one-sided anti-inflammatory. 47 , 50 , 51  This observation is discordant with findings\nfrom in vivo studies, where treatment with EP2 antagonist decreased\nthe levels of IL-1β, IL-6, and CCl2 and TNF in 3 models of SE\nin two species (rat and mice). 2 , 14 − 17 , 21  The in vitro up-regulation of\nTNF by EP2 antagonism can be explained by the finding that EP2 driven\ncAMP inhibits the release of TNF from the microglia or BV2 cells;\ntherefore EP2 agonist decreases the levels of TNF, whereas EP2 antagonist\ninterrupts this TNF blocking effect by the EP2/cAMP resulting in release\nof it, therefore we see EP2 antagonism increasing the TNF. 52 , 53  However, this mechanism is occurring only in cell culture studies\nin vitro. In vivo (in SE model studies), EP2 typically increases the\nTNF levels and EP2 antagonist decreases them, consistent with anti-inflammatory\nproperties for EP2 antagonist.\nMoreover, we recently reported\nthat EP2 is induced in rat microglia\nupon insult with LPS/IL-13, and the prolonged activation of EP2 with\nan agonist induces rat microglia death which can be prevented by treatment\nwith EP2 antagonists. 54  Cell death mediated\nby EP2 involves the activation of caspase-1 and -3 as well as generation\nof reactive oxygen species (ROS) promoting either pyroptosis or apoptosis\nmechanisms. In this study, microglia upon LPS/IL-13 treatment become\nswollen and round compared with the elongated form in resting condition,\nthe activation of microglia EP2 with butaprost change the morphology\nof microglia causing them to shrink suggesting apoptosis. Whether\nEP2 receptor activation really causes microglia death in vivo is not\nclear. But in several studies, 10 , 55  the inflammatory states\nof microglia are modulated by the presence and activation of EP2 receptor\ntoward a maladaptive immune state. Therefore, these in vitro results\nmust be interpreted individually and must not be compared with in\nvivo results to draw conclusions.\nGlobal EP2 deletion is detrimental\nto mice. One study showed that\nEP2 deletion decreases reproduction rate, reduces litter size, and\nsignificantly elevates blood pressure in mice when they are on a high-salt\ndiet compared with regular diet. 56  Another\nstudy reported that EP2 global deletion causes salt-sensitive hypertension\nand reduced fertility. 57  These two studies,\npublished 20 years ago, had created a stumbling block for the development\nof EP2 antagonists that now show beneficial effects in several animal\ndisease models as described in the  Strengths  section. These results obviously raise potential weaknesses, and\nthey could even be threats if they are replicated by pharmacological\nEP2 antagonism. To address these weaknesses, we have recently conducted\na study using two different EP2 antagonists, TG6-10-1 and TG11-77\n( Figure  2 ). TG6-10-1\nwas administered by acute dosing in mice and rats, and TG11-77 was\ndosed by chronic oral dosing via drinking water. The rodents were\nsubjected to regular or high-salt diets when they were on treatment\nwith these two EP2 antagonists. We measured the systolic and diastolic\nblood pressure, heart rate, and respiratory function in mice and rats.\nRegardless of the diet in mice, these two antagonists did not cause\nany of the adverse phenotypes that were found in mice with EP2 global\ndeletion. 58  The discordance of the results\nfrom EP2 gene knockout and pharmacological antagonism can be attributed\nto the role of EP2 in development at prenatal and postnatal stages.\nMoreover, in the adult stage, the EP2 receptor seems to carry a majority\nof COX2/PGE 2  driven inflammatory signaling. Therefore,\nthe adverse phenotypes found in EP2 deleted mice must be interpreted\nas weaknesses rather than real threats for advancing the EP2 receptor\nas a therapeutic target with small molecule EP2 antagonists.\nIt has been very well-known that PGE 2 , via G protein-coupled\nsignaling, is involved not only in inflammation but also in bone-formation\nand bone-healing, embryo implantation, induction of labor, and vasodilation\nindicating a “yin–yang” nature of PGE 2  signaling depending on the injury and the disease. 2  Endogenous PGE 2  expression increased after bone\nfractures, and administration of PGE 2  also stimulated bone\nformation in animal models. 59 − 61  Both EP2 and EP4 receptors expressed\nin bone cells and marrow stromal cells are shown to play an important\nrole in bone formation and resorption 62  determined by using mice with either EP2 or EP4 knockout and the\nselective agonists of these two receptors. A selective EP2 agonist\nCP-533,536 ( Figure  3 ) directly injected into bone marrow healed the fractured bone in\nrat and canine models. 6 , 63  Pfizer has promoted this agonist\nfor human clinical trials to examine efficacy, safety, and tolerability\nin subjects with closed fracture of the tibial shaft ( https://clinicaltrials.gov/ct2/show/NCT00533377 ). Although clinical study results are not published to conclude\nthe clinical utility of this agonist and clinical proof-of-evidence\nfor EP2 agonism for fractured bones, the in vivo results from multiple\nmodels provide a strong rationale for EP2 agonists for local bone\naugmentation, bone repair, and healing. 6 , 63  Similar beneficial\neffects were also found with use of selective EP4 agonists in these\nbone-fracture and bone-repair models, 64 , 65  suggesting\nboth receptors are involved in the bone repair and healing process.\nSimilarly, Pfizer also promoted an EP2 agonist CP-544,326 (PF-04217329,\naka, taprenepag isopropyl) ( Figure  3 ) for the treatment of open-angle glaucoma and ocular\nhypertension ( www.clinicaltrials.gov ). In light of these findings, it is reasonable and important to\nquestion whether an acute or chronic treatment of EP2 antagonist would\ncompromise healthy bones and weaken them. To address this question,\nwe recently conducted a study with EP2 antagonist TG11-77. Upon chronic\ndosing of (134 mg/kg/day free base) TG11-77·HCl to mice in drinking\nwater for 28 days, the tibia and femur from hind limbs were analyzed\nfor bone mass through diaphyseal scan and trabecular network through\nmetaphyseal scan by microcomputed tomography (μCT). 58  Overall, this study showed that EP2 antagonist\ntreatment has no adverse effect on bone volume and density in healthy\nmice. These results dampen the potential threat to healthy bones and\nstrengthen the advancement of EP2 antagonist for clinical use.\n\nInflammation is an ongoing feature found\nin several central nervous\nsystem and peripheral diseases. In general, inflammation affects >100\nmillion people in the USA. The global anti-inflammatory market is\nprojected to reach $135 billion by 2027 with 4.8% compound annual\ngrowth rate (CAGR). Nonsteroidal anti-inflammatory drugs (NSAIDs)\nand cyclooxygenase-2 (COX2) inhibitors looked promising, but they\nwere limited by their gastrointestinal and cardiovascular toxicity.\nIn many inflammatory conditions, induction of COX2 and mPGES-1 was\nobserved, together leading to synthesis of PGE 2  driving\ndownstream signaling through EP2 and EP4 receptors by synthesis of\ncAMP, EP1 by immobilization of intracellular Ca 2+ , or EP3\nvia inhibiting cAMP. Among these four PGE2 receptors, EP2 seems to\nact as an inflammatory mediator in the majority of in vivo studies\nexamined so far (see  Strengths  section),\nwhereas EP4 can act as a proinflammatory or an anti-inflammatory agent\ndepending on the disease context. For example, EP4 receptor acts as\na proinflammatory agent in rheumatoid and osteoarthritis conditions 66 , 67  but as an anti-inflammatory agent in cardiovascular and Alzheimer’s\ndisease models. 68 , 69  The role of EP1 and EP3 seems\nlimited in terms of promoting inflammation. Therefore, EP2 provides\na tremendous opportunity to develop targeted therapeutics that should\nbypass the adverse cardiovascular events found with the use of COX2\ninhibitors rofecoxib (Vioxx) and valdecoxib (Bextra). 70 , 71\nSimilarly, cancer impacts 18 million people worldwide, leading\nto about 10 million deaths a year. The global oncology market was\nUS $286 billion in the year 2021, which is expected to reach $581\nbillion by 2030, with CAGR 8.2% from 2022 to 2030. There are many\ncancer subtypes impacting various segments of the population. The\nproliferation, tumor growth, and metastasis of many of these cancers\nis associated with the inflammatory tumor microenvironment. Interestingly,\nCOX2, PGE 2  and EP2 all are driving this malignant tumor\ngrowth; therefore, selectively targeting EP2 receptor should offer\ntherapeutic advantages that are not found with the use of generic\nCOX2 inhibitors and drugs with other mechanisms of action. One expects\nthat targeting the EP2 receptor selectively downstream of complex\nsignaling by COX2 should spare the physiologically relevant cardioprotective\nprostanoid receptor IP, which is activated by COX2 derived PGI 2 , and platelet modulator TP receptor, which is activated by\nCOX2 derived TXA 2  ligand. 72 , 73\nThere\nare several medically unaddressed diseases for which treatments\nare urgently needed. Capturing the impacts and unmet needs of each\ndisease is beyond the scope of this Perspective. Just to give an example,\nAlzheimer’s disease (AD), characterized by the onset of cognitive\nimpairment, is the most common cause of dementia. It affects 6 million\npeople in the USA, and this number is expected to grow to 14 million\nby 2050. According to a recent review (by Kim et al.), 74  there were about 543 interventional clinical\ntrials among the total of 2695 clinical trials conducted for AD between\n2004 and 2021. Among these, 41% failed in phase III trials and 59%\nfailed in Phase II. These trials included 64% disease modifying and\n36% symptomatic agents. Nonetheless, the FDA approved a monoclonal\nantibody (drug) in 2021 (Biogen/Esai’s aducanumab; aka., Aduhelm),\ndespite unanimous recommendations by the scientific review committee\nto reject the approval. 74  This year (January\n6, 2023), another antibody named lecanemab-irmb (Leqembi) was approved\nthrough the accelerated approval pathway by the FDA for the treatment\nof AD ( www.leqembi.com ).\nDue to the paucity of success in drug discovery and development against\nAD, exploring novel proof-of-concept drugs that work through a novel\nbiological target such as EP2 receptor seems an important task for\ninvestigation. Moreover, due to known adverse cardiovascular events\nwith chronic use of COX2 drugs, 70 − 73  there is little to no incentive or enthusiasm to\nconduct additional long-term clinical trials with COX2 drugs for debilitating\ndiseases such as AD, 75  post-traumatic epilepsy,\nor other chronic neurodegenerative diseases. Thus, targeting EP2 receptors\nwith small molecules provides enormous opportunities for clinical\ndevelopment.\n\nThe real threats\nfor targeting EP2 receptor by a pharmacological\napproach are elusive except for use against endometriosis. EP2 expression\nis strong in luminal epithelium at the implantation sites and may\nserve as a marker for uterine receptivity suggesting its role in embryo\nimplantation in mouse and rat. 76 , 77  Because endometriosis\nimpacts women at childbearing age, this potential threat must be addressed\nwith pharmacological antagonism with a specific EP2 antagonist, because\nEP4 and IP receptors are also highly expressed at the sites of embryo\nimplantation in uterus, and they may play a compensatory role for\nEP2 in this context.\nEP2 promotes cellular signaling cascades\nvia several intracellular\nmolecules and pathways. As discussed above, it mediates Gα s -dependent cAMP driven PKA and Epac signaling cascades on\none side, which drive inflammation, neurodegeneration, and neuronal\nplasticity, and G-protein independent signaling via β-arrestin\nsignaling on the other that drives cancer proliferation and metastasis\nand tumor development. Moreover, the anabolic activity of EP2 in the\nbone and bone marrow is also coupled to cAMP mediated signaling. All\nof these could give mixed conclusions to drug discovery and pharmaceutical\ncommunities and limited clarity on the therapeutic indication for\nwhich the advancement of EP2 drugs could be prioritized.\nSo\nfar, there is one EP2 targeted drug, omidenepag isopropyl (aka,\nOmlonti) ( Figure  3 ),\nis clinically approved by the FDA for the reduction of elevated intraocular\npressure in patients with primary open-angle glaucoma or ocular hypertension\n( https://www.omlonti.com ). A trailing second candidate in the class, PF-0417329 (prodrug\nof CP-544,326, see  Figure  3 ), also underwent clinical evaluation in humans ( NCT00934089 )\nand it significantly reduced intraocular pressure in primary open-angle\nglaucoma and ocular hypertension ( NCT00572455 ). 78\nAs to the EP2 antagonists, one candidate EP2 antagonist,\nPF-04418948,\nwent through Phase 1 human clinical trials examining the safety and\ntolerability of the compound by single and escalating doses ( https://beta.clinicaltrials.gov/study/NCT01002963 ). The data seem compelling and showed dose-linear increase in AUC\nfrom 30 mg/kg to 1000 mg/kg (but not beyond) doses, and treatment\nwas well tolerated with no cardiovascular events or renal toxicity\n(measured by KIM-1 molecule); however, it showed mild hyperbilirubinemia\n(dose-dependent increase in bilirubin) which is associated with its\nstrong inhibition activity against blood transporter OATP1B1. 79  Since then, Pfizer has made some organizational\nchanges, and as a result the subsequent development of this project\nhas been terminated (personal communication). Nonetheless, PF-04418948\nis very selective to the EP2 receptor, and it was able to reverse\nthe PGE 2  induced relaxation of mouse trachea at IC 50  = 2.7 nM; it suppressed butaprost induced cutaneous blood\nflow by oral-dosing at 3 mg/kg in rat. 32  PF-04418948 is a carboxylic acid derivative and displayed low volume\nof distribution and clearance with terminal plasma half-life of 8.8\nh with oral bioavailability of 78%; 32  however,\nit is brain-impermeable, and therefore, it can be used for blocking\nperipheral EP2 effects. Likewise, Amgen has investigated EP2 as a\ntarget for drug discovery and identified a lead candidate from high-throughput\nscreening (HTS) and SAR studies. The lead candidate molecule C52 seems\nto be a selective antagonist of EP2 over other prostanoid receptors,\nand it is highly brain-permeable (B/P ratio 0.7–0.9) with a\nplasma half-life of 3.4 h and oral bioavailability of 44%. 12  However, Amgen did not pursue this project further\nfor strategic reasons, and they closed the research site where this\nprogram evolved, and the program was also terminated (personal communications).\nOur laboratory has made significant contributions in the creation\nand development of a novel class of EP2 antagonists. The first-generation\nresearch lead compound in the class is TG6-10-1, 14  which has some structural weaknesses. It possesses an acryl\namide moiety, which potentially acts as a Michael acceptor for a variety\nof proteins and amino acids to form adducts in biological systems,\nwhich could pose some limitations for clinical development. The second\nresearch lead candidate in the program was TG8-260, which is highly\npotent and orally bioavailable but is not brain-permeable (B/P ratio\n0.04). Moreover, it has shown very potent cytochrome P450 (CYP) inhibition\nactivity against several CYP450 enzymes; 51  therefore, it has a potential limitation of displaying drug–drug\ninteractions. Very recently, we have reported the preclinical characterization\nof current lead molecule TG11-77, which has passed several IND-related\nADME-PK tests. 21  In comparison to Pfizer\ncompound PF-04418948 and Amgen compound C52, it has shown weak inhibition\nactivity against blood transporters (unpublished), and it is currently\ngoing through additional pre-IND requisite dose–response toxicokinetic\ntests in dogs.\n\nSo far, clinical proof-of-concept\n(POC) with EP2 agonist omidenepag\nisopropyl was achieved for ophthalmic use to conclude that EP2 is\na druggable target. However, clinical POC with an EP2 antagonist is\nyet to be achieved for any indication. Nonetheless, based on the substantial\nin vivo efficacy data from SE models in our laboratory, where EP2\nantagonism was proven to be anti-inflammatory in three chemically\ninduced models of SE (pilocarpine, kainate, diisopropyl-fluorophosphate\n(DFP)) in two rodent species ( Table  2 ) and where the anti-inflammatory effect of EP2 antagonism\nhas been translated into cognitive/memory improvements in those models,\nit is important to advance a clinical candidate toward attenuation\nor delay of the cognitive impairments in SE patients, and other patients\n(such as those suffering postoperative surgery or severe infections\nby RSV and SARS viruses) that are prone to develop cognitive impairments\nand also patients with autoimmune disorders like multiple sclerosis.\nDepending on the disease, the EP2 antagonist can be administered as\na first line monotherapy (for cancer and other peripheral inflammatory\ndiseases) or an adjunctive therapy along with first line antiseizure\ndrugs for the SE indication. For the treatment of AD, it is crucial\nto identify the right time to begin the treatment and the duration\nof the treatment with a novel anti-inflammatory agent. These studies\nmust be guided by the current understanding and the trajectory of\nmicroglial activation and its house-keeping performance (Aβ-clearance)\nduring the course of development of Alzheimer’s disease. 80  For the treatment of brain cancers, the studies\nmust develop a brain-permeable candidate with requisite pharmacokinetics\nand acceptable ADMET properties that facilitate Q.I.D. or B.I.D. dosing\nin patients. These goals are all achievable in the foreseeable feature.\nIn the efficacy studies, rodents\nwere exposed to listed EP2 antagonists for a short period of time,\n2–30 h, following 1 h SE. TG8-260 is a brain-impermeable compound,\nwhereas TG11-77 is brain permeable, and TG6-10-1 has excellent brain-permeability.\nDetails are provided in  Table  1  and the references cited in  Table  1 . Structures are shown in  Figure  2 .\nNot determined.\nThe EP2 receptor plays “yin–yang”\nphysiological and pathological roles. The advent of selective EP2 antagonists and agonists\ncontributed significantly to the conclusion that EP2 promotes inflammation\nin CNS diseases and cancer. With the\nFDA approval of omidenepag for glaucoma and\nsound beneficial effects of EP2 antagonism in status epilepticus and\ncancer models, the EP2 receptor seems to be a novel druggable target. This Perspective summarizes the pros and\ncons of targeting\nEP2 receptors with pharmacological agents.\nThe EP2 receptor plays “yin–yang”\nphysiological and pathological roles.\nThe advent of selective EP2 antagonists and agonists\ncontributed significantly to the conclusion that EP2 promotes inflammation\nin CNS diseases and cancer.\nWith the\nFDA approval of omidenepag for glaucoma and\nsound beneficial effects of EP2 antagonism in status epilepticus and\ncancer models, the EP2 receptor seems to be a novel druggable target.\nThis Perspective summarizes the pros and\ncons of targeting\nEP2 receptors with pharmacological agents.","source_license":"CC-BY-4.0","license_restricted":false}